{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T15:54:14Z","timestamp":1777046054238,"version":"3.51.4"},"reference-count":61,"publisher":"MDPI AG","issue":"5","license":[{"start":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T00:00:00Z","timestamp":1776988800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Algorithms"],"abstract":"<jats:p>(1) Background: Advances in machine learning (ML)-based survival modeling enable the analysis of high-dimensional biomedical data. However, many approaches rely on the proportional hazards (PH) assumption, which is frequently violated in oncology and can limit the interpretability of hazard ratio-based results. Using Estrogen Receptor (ER) status in the METABRIC breast cancer cohort as a case study, we propose a framework that integrates machine learning survival models with Restricted Mean Survival Time (RMST) to provide a more robust and clinically interpretable approach for survival analysis under non-proportional hazards. (2) Methods: Overall survival was analyzed in 1104 patients. PH violations were confirmed using Schoenfeld residuals and Kaplan\u2013Meier inspection. We compared four models: stratified Cox Elastic Net (Cox E-Net), Random Survival Forest (RSF), Gradient Boosting Survival Analysis (GBSA), and DeepHit. Performance was assessed using Harrell\u2019s C-index, time-dependent IPCW C-index, and Integrated Brier Score (IBS). RMST at 180 months was utilized to quantify absolute survival differences between ER subgroups. To improve the stability of the estimates, 200 bootstrap resamples were performed, and 95% confidence intervals were derived from the bootstrap distribution. (3) ER status demonstrated significant PH violation (p &lt; 0.005) with crossing survival curves. Discrimination (C-index 0.664\u20130.725) and calibration (IBS 0.149\u20130.169) were comparable across models, with RSF achieving the highest overall performance. Despite similar accuracy, survival curve structures differed substantially. Cox E-Net and RSF reproduced the observed crossing pattern, whereas GBSA generated smoother trajectories and DeepHit showed marked compression of subgroup separation. In the independent test cohort, the empirical RMST difference at 180 months was 16.6 months (ER-positive: 130.4; ER-negative: 113.8). Model-based RMST differences ranged from 1 month (DeepHit) to 27 months (Cox E-Net), with RSF and GBSA (12.8 and 13.8 months) most closely approximating the empirical benchmark. (4) Conclusions: We propose a novel, model-agnostic ML + RMST framework that addresses non-proportional hazards while providing quantifiable, time-specific clinical benefit. Moreover, models with similar discrimination and calibration produced markedly different survival curve behavior and absolute RMST estimates, demonstrating that accuracy metrics alone are insufficient for clinical interpretation. By linking prognostic modeling with absolute survival quantification, this framework advances survival evaluation beyond relative risk ranking toward individualized, clinically meaningful decision support.<\/jats:p>","DOI":"10.3390\/a19050329","type":"journal-article","created":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T14:28:23Z","timestamp":1777040903000},"page":"329","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["An RMST-Integrated Machine Learning Framework for Interpretable Survival Analysis Under Non-Proportional Hazards: Application to the METABRIC Cohort"],"prefix":"10.3390","volume":"19","author":[{"given":"Fangya","family":"Tan","sequence":"first","affiliation":[{"name":"Department of Analytics, Harrisburg University of Science and Technology, Harrisburg, PA 17101, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0142-1507","authenticated-orcid":false,"given":"Yang","family":"Zhou","sequence":"additional","affiliation":[{"name":"Department of Physics, The Graduate Center, City University of New York, New York, NY 10017, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shuqiao","family":"Li","sequence":"additional","affiliation":[{"name":"Department of Analytics, Harrisburg University of Science and Technology, Harrisburg, PA 17101, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Chun","family":"Jiang","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Rutgers, The State University of New Jersery, New Brunswick, NJ 08854, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5021-3739","authenticated-orcid":false,"given":"Jian-Guo","family":"Zhou","sequence":"additional","affiliation":[{"name":"Department of Oncology, The Second Affiliated Hospital of Zunyi Medical University, Zunyi 563000, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Srikar","family":"Bellur","sequence":"additional","affiliation":[{"name":"Department of Analytics, Harrisburg University of Science and Technology, Harrisburg, PA 17101, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,4,24]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"145","DOI":"10.4103\/2229-3485.86872","article-title":"Survival analysis in clinical trials: Basics and must know areas","volume":"2","author":"Singh","year":"2011","journal-title":"Perspect. Clin. Res."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"552","DOI":"10.1016\/S1470-2045(13)70093-7","article-title":"Axitinib versus sorafenib as second-line treatment for advanced renal cell carcinoma: Overall survival analysis and updated results from a randomised phase 3 trial","volume":"14","author":"Motzer","year":"2013","journal-title":"Lancet Oncol."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"222","DOI":"10.1200\/JCO.2009.22.7991","article-title":"Survival analysis of cancer risk reduction strategies for BRCA1\/2 mutation carriers","volume":"28","author":"Kurian","year":"2010","journal-title":"J. Clin. Oncol."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Ahmad, T., Munir, A., Bhatti, S.H., Aftab, M., and Raza, M.A. (2017). Survival analysis of heart failure patients: A case study. PLoS ONE, 12.","DOI":"10.1371\/journal.pone.0181001"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"13","DOI":"10.1002\/(SICI)1097-0258(20000115)19:1<13::AID-SIM279>3.0.CO;2-5","article-title":"Survival analysis for recurrent event data: An application to childhood infectious diseases","volume":"19","author":"Kelly","year":"2000","journal-title":"Stat. Med."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"1074","DOI":"10.1080\/01621459.1989.10478874","article-title":"The robust inference for the Cox proportional hazards model","volume":"84","author":"Lin","year":"1989","journal-title":"J. Am. Stat. Assoc."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"705","DOI":"10.1038\/ki.2008.294","article-title":"The analysis of survival data in nephrology: Basic concepts and methods of Cox regression","volume":"74","author":"Jager","year":"2008","journal-title":"Kidney Int."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Therneau, T.M., and Grambsch, P.M. (2000). Modeling Survival Data: Extending the Cox Model, Springer.","DOI":"10.1007\/978-1-4757-3294-8"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"431","DOI":"10.1038\/sj.bjc.6601119","article-title":"Survival analysis part II: Multivariate data analysis\u2013an introduction to concepts and methods","volume":"89","author":"Bradburn","year":"2003","journal-title":"Br. J. Cancer"},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Kleinbaum, D.G., and Klein, M. (1996). Survival Analysis a Self-Learning Text, Springer.","DOI":"10.2307\/2532873"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"1302811","DOI":"10.1155\/2021\/1302811","article-title":"Methods to analyze time-to-event data: The Cox regression analysis","volume":"2021","author":"Leonardis","year":"2021","journal-title":"Oxidative Med. Cell. Longev."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Collett, D. (2023). Modelling Survival Data in Medical Research, Chapman and Hall\/CRC.","DOI":"10.1201\/9781003282525"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"178","DOI":"10.1038\/nrclinonc.2011.217","article-title":"Hazard ratios in cancer clinical trials\u2014A primer","volume":"9","author":"Blagoev","year":"2012","journal-title":"Nat. Rev. Clin. Oncol."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"227","DOI":"10.1023\/A:1006133418245","article-title":"Time-dependence of hazard ratios for prognostic factors in primary breast cancer","volume":"52","author":"Hilsenbeck","year":"1998","journal-title":"Breast Cancer Res. Treat."},{"key":"ref_15","first-page":"176","article-title":"Regression models and non-proportional hazards in the analysis of breast cancer survival","volume":"33","author":"Gore","year":"1984","journal-title":"J. R. Stat. Soc. Ser. C (Appl. Stat.)"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"e9204","DOI":"10.1097\/MD.0000000000009204","article-title":"Reporting and methodological quality of survival analysis in articles published in Chinese oncology journals","volume":"96","author":"Zhu","year":"2017","journal-title":"Medicine"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1111\/j.2517-6161.1972.tb00899.x","article-title":"Regression models and life-tables","volume":"34","author":"Cox","year":"1972","journal-title":"J. R. Stat. Soc. Ser. B (Methodol.)"},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"c83","DOI":"10.1159\/000324758","article-title":"Survival analysis I: The Kaplan-Meier method","volume":"119","author":"Stel","year":"2011","journal-title":"Nephron Clin. Pract."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"182","DOI":"10.4097\/kja.d.18.00067","article-title":"Survival analysis: Part I\u2014Analysis of time-to-event","volume":"71","author":"In","year":"2018","journal-title":"Korean J. Anesthesiol."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"331","DOI":"10.1016\/j.otohns.2010.05.007","article-title":"A practical guide to understanding Kaplan-Meier curves","volume":"143","author":"Rich","year":"2010","journal-title":"Otolaryngol.\u2014Head Neck Surg."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Therneau, T.M. (1997). Extending the Cox model. Proceedings of the First Seattle Symposium in Biostatistics: Survival Analysis, Springer.","DOI":"10.1007\/978-1-4684-6316-3_5"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"499","DOI":"10.2307\/2531021","article-title":"Sample-size formula for the proportional-hazards regression model","volume":"39","author":"Schoenfeld","year":"1983","journal-title":"Biometrics"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"145","DOI":"10.1146\/annurev.publhealth.20.1.145","article-title":"Time-dependent covariates in the Cox proportional-hazards regression model","volume":"20","author":"Fisher","year":"1999","journal-title":"Annu. Rev. Public Health"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"551","DOI":"10.1002\/sim.4780080504","article-title":"Flexible regression models with cubic splines","volume":"8","author":"Durrleman","year":"1989","journal-title":"Stat. Med."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Gelman, A., and Hill, J. (2007). Data Analysis Using Regression and Multilevel\/Hierarchical Models, Cambridge University Press.","DOI":"10.32614\/CRAN.package.arm"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Royston, P., and Parmar, M.K. (2013). Restricted mean survival time: An alternative to the hazard ratio for the design and analysis of randomized trials with a time-to-event outcome. BMC Med. Res. Methodol., 13.","DOI":"10.1186\/1471-2288-13-152"},{"key":"ref_27","unstructured":"Hua, K. (2024). Design and Analysis of Clinical Trials with Restricted Mean Survival Time. [Doctoral Dissertation, Duke University]."},{"key":"ref_28","unstructured":"US Food and Drug Administration (2022). Acute Myeloid Leukemia: Developing Drugs and Biological Products for Treatment-Guidance for Industry. Guidance for Industry. Final."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"215","DOI":"10.1111\/biom.12384","article-title":"On the restricted mean survival time curve in survival analysis","volume":"72","author":"Zhao","year":"2016","journal-title":"Biometrics"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Calkins, K.L., Canan, C.E., Moore, R.D., Lesko, C.R., and Lau, B. (2018). An application of restricted mean survival time in a competing risks setting: Comparing time to ART initiation by injection drug use. BMC Med. Res. Methodol., 18.","DOI":"10.1186\/s12874-018-0484-z"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"1179","DOI":"10.1001\/jamacardio.2017.2922","article-title":"Restricted mean survival time as a measure to interpret clinical trial results","volume":"2","author":"Kim","year":"2017","journal-title":"JAMA Cardiol."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"495","DOI":"10.3348\/kjr.2022.0061","article-title":"Restricted mean survival time for survival analysis: A quick guide for clinical researchers","volume":"23","author":"Han","year":"2022","journal-title":"Korean J. Radiol."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"202","DOI":"10.1002\/pst.1846","article-title":"Comparison of the restricted mean survival time with the hazard ratio in superiority trials with a time-to-event end point","volume":"17","author":"Huang","year":"2018","journal-title":"Pharm. Stat."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"335","DOI":"10.1007\/s10985-004-4771-0","article-title":"Regression analysis of restricted mean survival time based on pseudo-observations","volume":"10","author":"Andersen","year":"2004","journal-title":"Lifetime Data Anal."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"1189","DOI":"10.1214\/aos\/1013203451","article-title":"Greedy function approximation: A gradient boosting machine","volume":"29","author":"Friedman","year":"2001","journal-title":"Ann. Stat."},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Imani, F., Chen, R., Tucker, C., and Yang, H. (2019). Random forest modeling for survival analysis of cancer recurrences. Proceedings of the 2019 IEEE 15th International Conference on Automation Science and Engineering (CASE), Vancouver, BC, Canada, 22\u201326 August 2019, IEEE.","DOI":"10.1109\/COASE.2019.8843271"},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Chen, T., and Guestrin, C. (2016, January 13\u201317). Xgboost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA.","DOI":"10.1145\/2939672.2939785"},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"355","DOI":"10.1093\/biostatistics\/kxj011","article-title":"Survival ensembles","volume":"7","author":"Hothorn","year":"2006","journal-title":"Biostatistics"},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Katzman, J.L., Shaham, U., Cloninger, A., Bates, J., Jiang, T., and Kluger, Y. (2018). DeepSurv: Personalized treatment recommender system using a Cox proportional hazards deep neural network. BMC Med. Res. Methodol., 18.","DOI":"10.1186\/s12874-018-0482-1"},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Lee, C., Zame, W., Yoon, J., and Van Der Schaar, M. (2018, January 2\u20137). Deephit: A deep learning approach to survival analysis with competing risks. Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, LA, USA.","DOI":"10.1609\/aaai.v32i1.11842"},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"2109","DOI":"10.1002\/sim.1802","article-title":"Overall C as a measure of discrimination in survival analysis: Model specific population value and confidence interval estimation","volume":"23","author":"Pencina","year":"2004","journal-title":"Stat. Med."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"1141","DOI":"10.1016\/0895-4356(91)90146-Z","article-title":"Assessing predictive accuracy: How to compare Brier scores","volume":"44","author":"Redelmeier","year":"1991","journal-title":"J. Clin. Epidemiol."},{"key":"ref_43","unstructured":"Ridgeway, G. (2026, January 01). Generalized Boosted Models: A Guide to the gbm Package; R Package Vignette: 2007; Volume 1. Available online: https:\/\/CRAN.R-project.org\/package=gbm."},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"e70031","DOI":"10.1002\/sim.70031","article-title":"Modeling the Restricted Mean Survival Time Using Pseudo-Value Random Forests","volume":"44","author":"Schenk","year":"2025","journal-title":"Stat. Med."},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3214306","article-title":"Machine learning for survival analysis: A survey","volume":"51","author":"Wang","year":"2019","journal-title":"ACM Comput. Surv. (CSUR)"},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"65","DOI":"10.1007\/s10462-023-10681-3","article-title":"Deep learning for survival analysis: A review","volume":"57","author":"Wiegrebe","year":"2024","journal-title":"Artif. Intell. Rev."},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"5672","DOI":"10.1093\/bioinformatics\/btaa1082","article-title":"Deep neural networks for predicting restricted mean survival times","volume":"36","author":"Zhao","year":"2020","journal-title":"Bioinformatics"},{"key":"ref_48","doi-asserted-by":"crossref","first-page":"27","DOI":"10.1023\/A:1020299707510","article-title":"Estrogen receptor breast cancer phenotypes in the Surveillance, Epidemiology, and End Results database","volume":"76","author":"Anderson","year":"2002","journal-title":"Breast Cancer Res. Treat."},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"325","DOI":"10.1080\/10408360600739218","article-title":"Estrogen receptors: Role in breast cancer","volume":"43","author":"Duffy","year":"2006","journal-title":"Crit. Rev. Clin. Lab. Sci."},{"key":"ref_50","doi-asserted-by":"crossref","first-page":"339","DOI":"10.1006\/scbi.2001.0389","article-title":"Estrogen receptor and breast cancer","volume":"Volume 11","author":"Sommer","year":"2001","journal-title":"Seminars in Cancer Biology"},{"key":"ref_51","doi-asserted-by":"crossref","unstructured":"Bilal, E., Dutkowski, J., Guinney, J., Jang, I.S., Logsdon, B.A., Pandey, G., Sauerwine, B.A., Shimoni, Y., Moen Vollan, H.K., and Mecham, B.H. (2013). Improving breast cancer survival analysis through competition-based multidimensional modeling. PLoS Comput. Biol., 9.","DOI":"10.1371\/journal.pcbi.1003047"},{"key":"ref_52","doi-asserted-by":"crossref","first-page":"346","DOI":"10.1038\/nature10983","article-title":"The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups","volume":"486","author":"Curtis","year":"2012","journal-title":"Nature"},{"key":"ref_53","doi-asserted-by":"crossref","first-page":"1167","DOI":"10.1136\/bmj.330.7501.1167","article-title":"Chemotherapy and hormonal treatments improve the 15 year survival rate for breast cancer","volume":"330","author":"Mayor","year":"2005","journal-title":"BMJ Br. Med. J."},{"key":"ref_54","doi-asserted-by":"crossref","first-page":"301","DOI":"10.1111\/j.1467-9868.2005.00503.x","article-title":"Regularization and variable selection via the elastic net","volume":"67","author":"Zou","year":"2005","journal-title":"J. R. Stat. Soc. Ser. B Stat. Methodol."},{"key":"ref_55","doi-asserted-by":"crossref","first-page":"1","DOI":"10.18637\/jss.v039.i05","article-title":"Regularization paths for Cox\u2019s proportional hazards model via coordinate descent","volume":"39","author":"Simon","year":"2011","journal-title":"J. Stat. Softw."},{"key":"ref_56","first-page":"1","article-title":"scikit-survival: A Library for Time-to-Event Analysis Built on Top of scikit-learn","volume":"21","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref_57","first-page":"2825","article-title":"Scikit-learn: Machine learning in Python","volume":"12","author":"Pedregosa","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"ref_58","doi-asserted-by":"crossref","first-page":"1317","DOI":"10.21105\/joss.01317","article-title":"lifelines: Survival analysis in Python","volume":"4","year":"2019","journal-title":"J. Open Source Softw."},{"key":"ref_59","doi-asserted-by":"crossref","first-page":"6968","DOI":"10.1038\/s41598-021-86327-7","article-title":"Explainable machine learning can outperform Cox regression predictions and provide insights in breast cancer survival","volume":"11","author":"Hendriks","year":"2021","journal-title":"Sci. Rep."},{"key":"ref_60","doi-asserted-by":"crossref","first-page":"873595","DOI":"10.1155\/2013\/873595","article-title":"A gradient boosting algorithm for survival analysis via direct optimization of concordance index","volume":"2013","author":"Chen","year":"2013","journal-title":"Comput. Math. Methods Med."},{"key":"ref_61","doi-asserted-by":"crossref","first-page":"bbaf521","DOI":"10.1093\/bib\/bbaf521","article-title":"SurvBoard: Standardized benchmarking for multi-omics cancer survival models","volume":"26","author":"Wissel","year":"2025","journal-title":"Brief. Bioinform."}],"container-title":["Algorithms"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1999-4893\/19\/5\/329\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T14:31:58Z","timestamp":1777041118000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1999-4893\/19\/5\/329"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,24]]},"references-count":61,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2026,5]]}},"alternative-id":["a19050329"],"URL":"https:\/\/doi.org\/10.3390\/a19050329","relation":{},"ISSN":["1999-4893"],"issn-type":[{"value":"1999-4893","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,24]]}}}